We have successfully pushed an update to all instances residing in the United States. For a summary of the release, please refer to the corresponding tag for Canadian zones.
We apologize for the unexpectedly lengthy upgrade process. Our root cause analysis yielded that a particular database query to retrieve transcription data with the new ASR model took much longer to execute than it did in our Canadian, European and staging instances. All instances should now be up and operating normally. We completed additional validations to ensure that normal performance has returned to all data processing sub-systems.